Leapfrog: Leapfrog

In jodie0399/NUTS: Hamiltonian Monte Carlo and No-U-Turn Sampler with Dual Averaging

Description

This function perform a leapfrog step. This function is a modified version of Leapfrog in the paper. It returns etra values: log posterior value and gradient of log posterior at the new position theta.tilde

This function performs a leapfrog step. This function is a modified version of Leapfrog in the paper. It returns extra values: log posterior value and gradient of log posterior at the new position theta.tilde

Usage

Leapfrog(theta, r, epsilon, L)

Leapfrog(theta, r, epsilon, L)

Arguments

theta	starting position
r	starting momentum
epsilon	step size
L	callable function: returns the value of log posterior and the gradient of log posterior prbability at given input
theta	starting position
r	starting momentum
epsilon	step size
L	callable function: returns the value of log posterior and the gradient of log posterior probability at given input

Value

the list of updated theta, r and the log posterior value at the updated point

the list of updated theta, r and the log posterior value at the updated point

jodie0399/NUTS documentation built on May 29, 2019, 1:06 a.m.